Synthetic Studies toward the Citrinadin A and B Core Architecture

Sep 19, 2013 - United States [email protected]. Received July 31, 2013. ABSTRACT. The core architecture of the citrinadin alkaloids has been ...
0 downloads 0 Views 328KB Size
ORGANIC LETTERS

Synthetic Studies toward the Citrinadin A and B Core Architecture

XXXX Vol. XX, No. XX 000–000

Devon A. Mundal and Richmond Sarpong* Department of Chemistry, University of California, Berkeley, California 94720, United States [email protected] Received July 31, 2013

ABSTRACT

The core architecture of the citrinadin alkaloids has been prepared in racemic form by utilizing a strategy that exploits the alkylation of 2-methoxypyridines. An initially planned indolizidine to quinolizidine transformation to build the D/E rings was unsuccessful. Success was ultimately gained by a direct alkylation to establish the citrinadin core architecture.

Citrinadins A (1, Figure 1) and B (2) are oxindole alkaloid natural products that were isolated from Penicillium citrinum by Jun’ichi Kobayashi and co-workers in 2004 and 2005, respectively.1,2 Recently, highly elegant total syntheses of alkaloids 1 and 2 have been completed by the groups of Martin and Wood, respectively.3,4 These contributions have resulted in the revision of the relative stereochemistry of the citrinadins to the representations shown for 1 and 2 in Figure 1. These stereochemical revisions now bring the structures of the citrinadins in line with the related PF1270 spiro-oxindole natural products (A C, 3 5), which were isolated from Penicillum waksmanii (strain PF1270) by Kushida et al.5 Synthetic interest in the citrinadins has been driven by their unusual structures, which feature a substituted quinolizidine and a cyclopentane ring replete with four tetrasubstituted carbon atoms, two of which are quaternary. (1) Tsuda, M.; Kasai, Y.; Komatsu, K.; Sone, T.; Tanaka, M.; Mikami, Y.; Kobayashi, J. Org. Lett. 2004, 6, 3087. (2) Mugishima, T.; Tsuda, M.; Kasai, Y.; Ishiyama, H.; Fukushi, E.; Kawabata, J.; Watanabe, M.; Akao, K.; Kobayashi, J. J. Org. Chem. 2005, 70, 9430. (3) Bian, Z.; Marvin, C. C.; Martin, S. F. J. Am. Chem. Soc. 2013, 135, 10886. (4) Kong, K.; Enquist, J. A., Jr.; McCallum, M. E.; Smith, G. M.; Matsumaru, T.; Menhaji-Klotz, E.; Wood, J. L. J. Am. Chem. Soc. 2013, 135, 10890. (5) Kushida, N.; Watanabe, N.; Okuda, T.; Yokoyama, F.; Gyobu, Y.; Yaguchi, T. J. Antibiot. 2007, 60, 667.

Figure 1. Citrinadins and PF1270 alkaloids.

In addition to their challenging structural attributes, which may inspire new synthetic strategies and tactics, 1 and 2, as well as related spiro-oxindoles such as 3 5, possess notable biological activity (e.g., cytotoxicity against murine leukemia L1210 cells for 1 (IC50 = 6.2 μg/mL) and 2 (IC50 = 10 μg/mL), whereas 3 5 are potent agonists of rat and human histamine receptor H3). As such, the syntheses of 1 and 2 or 3 5 could offer opportunities for a more comprehensive examination of their anticancer activity and their effects on the central nervous system. This is especially significant given the limited availability of these materials from their natural sources. It is therefore not surprising that, in addition to the recently completed total 10.1021/ol402177a

r XXXX American Chemical Society

syntheses of 1 and 2, there have been several approaches to these compounds that are each characterized by a creative, unique strategy to address the core architecture of these molecules.6 In this communication, we report our own synthetic studies toward the citrinadins, which is prompted by the recent disclosures of the Martin and Wood syntheses3,4 of 1 and 2, respectively, with which our approach shares several strategic features. As outlined in Scheme 1, we envisioned the citrinadins, particularly citrinadin B (2), arising (as was achieved in the Wood synthesis of 2) from functionalization at C-7 of oxindole 6 (where X = Br or I). In one of the key transformations of the synthesis, quinolizidine 6 would arise from indolizidine 7 (discussed in more detail in Scheme 2 below). It was imagined that alkylation of the 2-methoxypyridine portion of 8 by the epoxide functional group would provide eventual access to indolizidine 7. 2-Methoxypyridines offer several strategic advantages in complex molecule synthesis, which we have exploited in the past in the syntheses of several complex alkaloid natural products.7 For example, they are excellent surrogates for piperidine groups where the basic nitrogen atom is in essence protected given the mitigated basicity of the methoxypyridine nitrogen. As a corollary of this reduced basicity/nucleophilicity of the 2-methoxypyridine nitrogen, alkylation of the 2-methoxypyridine group is not general, especially using electrophiles other than alkyl triflates or halides. Thus, the annulation strategy proposed herein (i.e., 8 to 7) would serve to extend the scope of annulation reactions of 2-methoxypyridines. Fused indole tricycle 8 would in turn arise from hydrazine 9 (where X = Br or I), ketoester 10 (enol form shown), and 2-methoxypicoline 11.

Scheme 1. Retrosynthetic Analysis of Citrinadin B

(6) (a) Pettersson, M.; Knueppel, D.; Martin, S. F. Org. Lett. 2007, 9, 4623. (b) McIver, A. L.; Deiters, A. Org. Lett. 2010, 12, 1288. (c) Chandler, B. D.; Roland, J. T.; Li, Y.; Sorensen, E. J. Org. Lett. 2010, 12, 2746. (d) Guerrero, C. A.; Sorensen, E. J. Org. Lett. 2011, 13, 5164. (e) Albertshofer, K.; Tan, B.; Barbas, C. F. Org. Lett. 2012, 14, 1834. (7) (a) Bisai, A.; West, S. P.; Sarpong, R. J. Am. Chem. Soc. 2008, 130, 7222. (b) Larson, K. K.; Sarpong, R. J. Am. Chem. Soc. 2009, 131, 13244. (c) Bisai, V.; Sarpong, R. Org. Lett. 2010, 12, 2551. (d) Murphy, R. A.; Sarpong, R. Org. Lett. 2012, 14, 632. (e) Newton, J. N.; Fischer, D. F.; Sarpong, R. Angew. Chem., Int. Ed. 2013, 52, 1726. B

Particularly intriguing in our proposed approach to 2 is the conversion of 7 to 6. At the time that we initiated our studies, no direct precedent existed for this particular transformation. As such, a plan to study this conversion on a model system (12, Scheme 2) was hatched. It was expected that appropriate activation of the primary hydroxyl group of 12 would lead to aziridinium intermediate 13, which would form upon engaging the tertiary amine group. At that stage, the introduction of an appropriate nucleophile would lead to quinolizidine 14 by a SN1-like or an asynchronous SN2-like process capable of delivering the requisite stereochemistry (either via substrate or reagent control). While we cannot exclude the possibility of a the direct SN2 opening of aziridinium intermediate 138b (which would lead to the stereochemistry opposite to that depicted in compound 14), our hypothesis is supported by the recent reports from the Wood group, where the conversion of 15 to 17 may pass through an intermediate (16) where the tertiary amine offers a level of anchimeric assistance.4,8

Scheme 2. Proposed Indolizidine to Quinolizidine Conversion via an Aziridinium Intermediate

Our synthesis of the model indolizidine compound 12 commenced with the preparation of 24 as outlined in Scheme 3. Commercially available 2,2-dimethylcyclohexane-1,3-dione (18)9 was converted to monoketal 19, which was subjected to a Claisen reaction to afford 10. A standard triflation of 10 followed by Negishi cross-coupling10 with freshly prepared 21 smoothly affords the expected adduct, which upon hydrolysis of the ketal group yields ketone 22. Fischer indolization11 of 22 via the intermediacy of hydrazone 23 affords dihydrocarbazole derivative 24 (following Boc protection of the indole nitrogen). Access to 24 set the stage for the synthesis of the indolizidone derivative 27 as outlined in Scheme 4. Reduction of ester 24 with DIBAL produces an allylic alcohol, (8) (a) The Wood group has also demonstrated this epoxide opening on C-7 substituted derivatives of 15. See: Smith, G. M. Progress toward the Total Synthesis of the Citrinadins, Ph.D. Thesis, Colorado State University, Fort Collins, CO, 2012. (b) The work of Wonjacynska et al. has shown that aziridinium openings may proceed with retention of stereochemistry, which supports the possible intermediacy of 16: Wojaczynska, E.; Turowska-Tryk, I.; Skarzewski, J. Tetrahedron 2012, 68, 7848–7854. (9) For a recent application of 18 in synthesis, see ref 3. (10) For an early account, see: Negishi, E. Acc. Chem. Res. 1982, 15, 340. (11) For a review on the Fischer indole synthesis, see: Martin, M. J.; Dorn, L. J.; Cook, J. M. Heterocycles 1993, 36, 157. Org. Lett., Vol. XX, No. XX, XXXX

Scheme 3. Synthesis of Dihydrocarbazole Derivative 24

which upon epoxidation gives epoxide 25. After a protracted optimization, conversion of 25 to 26 could be effected with MgCl2 as the Lewis acid and NaI to accomplish the requisite demethylation. At this stage, hydrogenation of the pyridine occurs with good levels of diastereoselectivity to afford pentacycle 27, the structure of which was unambiguously confirmed by single-crystal X-ray analysis.12

Scheme 4. Synthesis of Pentacycle 27

Scheme 5. Synthesis of Spiro-oxindole Indolizidine 12

model compound to study the proposed key indolizidine to quinolizidine conversion that would deliver the citrinadin skeleton (see 12 to 14, Scheme 2).13 Despite numerous attempts using a wide selection of conditions (Table 1) and the encouraging related precedent of Wood,4,14 we have so far not had success in accomplishing the conversion of 12 to 14. For example, treatment of 12 under Cossy’s conditions15 (entry 1) with trifluoroacetic anhydride and triethylamine results in the formation of an intermediate trifluoroacetate. However, quinolizidine formation did not occur, but instead, the starting material was recovered (presumably after base-promoted hydrolysis of the trifluoroacetate). Other conditions that combine hydroxyl activation followed by the introduction of a nucleophile simply result in the decomposition of 12. Thus, while it would appear (from the precedent of Wood) that a quinolizidine structural motif may result if an aziridinium intermediate is accessed, our studies thus far suggest that indolizidine 12 may not represent the ideal substrate to access the requisite aziridinium species.

Table 1. Attempted Indolizidine to Quinolizidine Conversion

The next task was to convert fused pentacycle 27 to spirocyclic oxindole 28 (Scheme 5), which is readily accomplished with high levels of diastereocontrol using in situ generated dimethyldioxirane (DMDO; from oxone and acetone). The excellent diastereoselectivity can in part be attributed to a convex face approach of the DMDO, which may also be directed to the R face of 27 by hydrogen bond interactions with the primary hydroxyl group at the C/D ring fusion. Sequential removal of the Boc and amide carbonyl groups in 28 gives 12, which is the requisite (12) CYLView depictions of the X-ray crystal structures are shown. Thermal ellipsoids shown at 50% probability. Most hydrogens removed for clarity. See the Supporting Information for more details. Org. Lett., Vol. XX, No. XX, XXXX

(13) For examples of prolinol to piperidine ring expansions via aziridinium intermediates, see: (a) Cossy, J.; Dumas, C.; Pardo, D. G. Eur. J. Org. Chem. 1999, 1693. (b) Cochi, A.; Pardo, D. G.; Cossy, J. Eur. J. Org. Chem. 2012, 2023. (c) Wojaczynska, E.; Turowska-Tyrk, I.; Skarzewski, J. Tetrahedron 2012, 68, 7848. (d) Abe, H.; Aoyagi, S.; Kibayashi, C. J. Am. Chem. Soc. 2005, 127, 1473. (e) Jarvis, S. B. D.; Charette, A. B. Org. Lett. 2011, 13, 3830. (14) Also see ref 8. (15) See refs 13a and 13b. C

Scheme 6. Synthesis of Spiro-oxindole 32

Despite the disappointing observations in the attempted conversion of 12 to 14, our synthetic studies have identified an alternative sequence to access the carbon skeleton of the citrinadins bearing a quinolizidone framework (Scheme 6). Thus, treatment of a solution of epoxy alcohol 25 with triflic anhydride in the presence of 2,6-di-tert-butylpyridine (2,6-DTBP) yields pyridinium salt 29. Aqueous workup of this salt followed by treatment with sodium iodide in refluxing acetonitrile gives pentacyclic pyridone 30 in 75% yield over the two steps. Catalytic hydrogenation of pyridone 30 proceeds with excellent diastereoselectivity (>20:1) to give epoxy quinolizidone 31 in 82% yield. At this stage, oxidation of indole 31 with dimethyldioxirane (generated in situ from oxone and acetone) gives spiro-oxindole 32, the structure

D

of which was unambiguously supported by single-crystal X-ray analysis (see CYLView in Scheme 6).12 Although the relative stereochemistry at C-3 and C-16 (citrinadin numbering) in 32 is as desired, the C-18 stereocenter will require inversion. Studies to effect the inversion of the C-18 stereochemistry as well as nucleophilic opening of the epoxide group in 32 and derivatives thereof at C-8 are the subject of our ongoing studies. In conclusion, we have applied a methoxypyridine alkylation strategy to the synthesis of the pentacyclic carbon skeleton of the citrinadin natural products. A planned indolizidine ring expansion/nucleophile trapping via an aziridinium intermediate has thus far not been successful despite encouraging literature precedent from the work of Wood et al. An alternative methoxypyridine alkylation has enabled access to the citrinadin pentacyclic core and sets the stage for future studies to construct the fully substituted pentacyclic core of the citrinadin natural products. Acknowledgment. We thank the NIH (NIGMS R01 086374 and F32CA167908-01) for financial support, and a Ruth L. Kirschstein NRSA postdoctoral fellowship to D.A.M. R.S. is a Camille Dreyfus Teacher-Scholar. We thank Dr. Antonio DiPasquale (UC Berkeley) for solving the crystal structures of 27, 28, and 32 and acknowledge the CYLView program (developed by Prof. Claude Y. Legault, Dept. of Chemistry, Universite Sherbrooke) for X-ray depictions. Supporting Information Available. Experimental details and copies of 1H and 13C spectra for all new compounds. This material is available free of charge via the Internet at http://pubs.acs.org. The authors declare no competing financial interest.

Org. Lett., Vol. XX, No. XX, XXXX